The salient features of the theory of electron-cyclotron absorption an
d emission for the parameter range for which the radiation spectrum ha
s a line structure are reviewed and contrasted with erroneous results
of a recent paper by Arunasalam [Phys. Fluids B 4, 1643 (1992)]. In pa
rticular, (i) the relativistic and nonrelativistic forms of the dielec
tric tensor relevant to describing the electron-cyclotron absorption a
re considered, their range of validity is displayed, and on this basis
an incorrect suggestion in the above paper on how to obtain the relat
ivistic dielectric tensor from the nonrelativistic one is refuted; (ii
) the crucial role of the thermal effects on the mode polarization, wh
ich in turn affects the absorption profile around the fundamental freq
uency, is emphasized, showing that using a ''Trubnikov 0-mode factor''
as advocated in the above paper is not correct; (iii) the profile of
the absorption around the fundamental frequency results to be single h
umped in all cases, in contrast to the assertion in the above paper th
at the absorption profile of the first harmonic 0 mode is double humpe
d; (iv) the radiation from a nonuniformly magnetized plasma is evaluat
ed for a bi-Maxwellian electron distribution showing, in particular, t
hat the nonrelativistic treatment used in the above paper is incorrect
for perpendicular propagation; (v) an expression for the local radiat
ion temperature obtained in the above paper and upon which the analysi
s of the emission spectra of supershot plasmas in the Tokamak Fusion T
est Reactor (TFTR) has been based by Taylor et al [Proceedings of the
Eighth Workshop on Electron Cyclotron Emission and Heating, Gut Ising,
Germany, 1992, (Max-Planck Institut fur Plasmaphysik, Garching, 1993)
, IPP III/186, Vol. 1, p. 277] is shown not to be valid.